Skip to main content
. 2010 Feb 10;26(7):860–866. doi: 10.1093/bioinformatics/btq049

Table 1.

Improvement in TF role prediction using comparative genomics

Escherichia coli K12
Saccharomyces cerevisiae
Homo sapiens
Single Multiple Increase Single Multiple Increase Single Multiple Increase
species species (%) species species (%) species species (%)
Significant TF–GO term pairs 0 14 NA 420 733 75 371 1112 200
GO terms per TF tested 0 0.16 NA 3.4 5.9 75 6.6 19.8 200
Covered TFs 0 9 NA 99 113 14 48 56 17
Term specificity 0 4.0 NA 4.5 4.6 2 3.8 4.2 11

TFs tested 85 124 56

The table shows FDR (q≤0.05) results for single- and multiple-species Gomo. The results shown are the total number of most-specific significant pairs (‘significant TF–GO term pairs’), the average number of most-specific GO terms per TF tested (‘GO terms per TF tested’), the number of TFs with at least one significant TF–GO term pair (‘covered TFs’), the average depth in the GO hierarchy of significant GO terms (‘term specificity’), and the total number of TFs in each experiment (‘TFs tested’). All results are for GC-compensated Ama scores and ‘full’ promoters of 500, 750 and 1000 bp for enterobacter, yeast and mammals, respectively. NA, not applicable.